Performance-Based Migration Among Data Storage Devices

1. A method, comprising: identifying, by a computer system, a difference between one or more performance metrics of an application and a service-level agreement (SLA) of the application; calculating, by the computer system, a numeric correlation between the one or more application performance metrics and one or more disk input/output (I/O) metrics representing the disk I/O performance of a data storage device used by the application; comparing the numeric correlation to a threshold value; determining a saturation of disk I/O by the application; determining that an SLA violation occurred based on the identified difference, the saturation of disk I/O by the application, and the numeric correlation exceeding the threshold value; and outputting a recommendation to migrate the application from the data storage device to a different type of data storage device.

2. The method of claim 1 , wherein determining the saturation of disk I/O by the application comprises: comparing the one or more disk I/O metrics with baseline values for the one or more disk I/O metrics on the data storage device; and identifying the saturation of disk I/O when a disk I/O metric in the one or more disk I/O metrics exceeds a corresponding baseline value.

3. The method of claim 2 , wherein determining the saturation of disk I/O by the application further comprises: for each disk I/O metric in the one or more disk I/O metrics: determining a maximum value of the disk I/O metric; and scaling the maximum value to obtain a baseline value for the disk I/O metric prior to comparing the disk I/O metric with the baseline value.

4. The method of claim 1 , wherein the correlation comprises a rank correlation.

5. The method of claim 1 , wherein the one or more performance metrics comprises at least one of: a latency; and a throughput.

6. The method of claim 2 , wherein the one or more disk I/O metrics comprise an I/O operations per second (IOPS).

7. The method of claim 2 , wherein the one or more disk I/O metrics comprise an I/O rate.

8. The method of claim 2 , wherein the one or more disk I/O metrics comprise an I/O latency.

9. The method of claim 1 , wherein the data storage device comprises a hard disk drive (HDD) and the different type of data storage device comprises a solid-state drive (SSD).

10. The method of claim 1 , wherein identifying the difference between the performance of the application and the SLA of the application comprises: identifying a violation of the SLA by the application over a pre-specified period.

11. The method of claim 2 , wherein determining the saturation of disk I/O by the application further comprises calculating the baseline values for the one or more disk I/O metrics on the data storage device by, for each of the one or more disk I/O metrics: identifying a maximum value for the disk I/O metric during processing of a benchmark workload; and dividing the maximum value by a numeric factor to obtain a value representing saturation of disk I/O processing of normal workloads on the data storage device; wherein the baseline value for the disk I/O metric comprises the obtained value.

12. An apparatus, comprising: one or more processors; and memory storing instructions that, when executed by the one or more processors, cause the apparatus to: identify a difference between one or more performance metrics of an application and a service-level agreement (SLA) of the application; calculate a numeric correlation between the one or more application performance metrics and one or more disk input/output (I/O) metrics representing the disk I/O performance of a data storage device used by the application; compare the numeric correlation to a threshold value; determine a saturation of disk I/O by the application; determine that an SLA violation occurred based on the identified difference, the saturation of disk I/O by the application, and the numeric correlation exceeding the threshold value; and output a recommendation to migrate the application from the data storage device to a different type of data storage device.

13. The apparatus of claim 12 , wherein determining the saturation of disk I/O by the application comprises: comparing the one or more disk I/O metrics with baseline values for the one or more disk I/O metrics on the first type of data storage device; and identifying the saturation of disk I/O when a disk I/O metric in the one or more disk I/O metrics exceeds a corresponding baseline value.

14. The apparatus of claim 13 , wherein the one or more disk I/O metrics comprise at least one of: an I/O operations per second (IOPS); an I/O rate; and an I/O latency.

15. The apparatus of claim 14 , wherein the one or more performance metrics comprises at least one of: a latency; and a throughput.

16. The apparatus of claim 13 , wherein determining the saturation of disk I/O by the application further comprises calculating the baseline values for the one or more disk I/O metrics on the data storage device by, for each of the one or more disk I/O metrics: identifying a maximum value for the disk I/O metric during processing of a benchmark workload; and dividing the maximum value by a numeric factor to obtain a value representing saturation of disk I/O processing of normal workloads on the data storage device; wherein the baseline value for the disk I/O metric comprises the obtained value.

17. A system, comprising: an analysis module comprising a non-transitory computer-readable medium comprising instructions that, when executed, cause the system to: identify a difference between one or more performance metrics of an application and a service-level agreement (SLA) of the application; calculate a numeric correlation between the one or more application performance metrics and one or more disk input/output (I/O) metrics representing the disk I/O performance of a data storage device used by the application; compare the numeric correlation to a threshold value; determine a saturation of disk I/O by the application; determine that an SLA violation occurred based on the identified difference, the saturation of disk I/O by the application, and the numeric correlation exceeding the threshold value; and a management module comprising a non-transitory computer-readable medium comprising instructions that, when executed, cause the system to output a recommendation to migrate the application from the data storage device to a different type of data storage device when the correlation exceeds a threshold.

18. The system of claim 17 , wherein determining the saturation of disk I/O by the application comprises: comparing the one or more disk I/O metrics with baseline values for the one or more disk I/O metrics on the data storage device; and identifying the saturation of disk I/O when a disk I/O metric in the one or more disk I/O metrics exceeds a corresponding baseline value.

19. The system of claim 18 , wherein determining the saturation of disk I/O by the application further comprises calculating the baseline values for the one or more disk I/O metrics on the data storage device by, for each of the one or more disk I/O metrics: identifying a maximum value for the disk I/O metric during processing of a benchmark workload; and dividing the maximum value by a numeric factor to obtain a value representing saturation of disk I/O processing of normal workloads on the data storage device; wherein the baseline value for the disk I/O metric comprises the obtained value.